56 RESULTS WITH BARLEY 



same quantities of manure, the return during the same forty 

 years has amounted to about 60 per cent of that theoretically 

 possible. How are we to explain this great difference ? We 

 must first remark that the apparent increase by ammonia in 

 the barley field is probably somewhat above the truth. 

 During this long continued experiment the yield on plots i o, 

 2 o, and 4 o, has considerably diminished ; that is to say, the 

 base line from which the increase by ammonia is calculated 

 has steadily fallen, and the apparent increase by ammonia has 

 consequently risen. This is best shown by a comparison 

 of the produce of plots 4 o and 4A, as in this case the results 

 are not affected by any exhaustion of ash constituents in the 

 course of the experiment. The increased produce plot of 2A 

 over plot 2 o, and of 4A over plot 4 o, when calculated as the 

 return given by i cwt. of sulphate of ammonia, has been in the 

 four successive periods of ten years as follows : 



RETURN FOR i CWT. OF SULPHATE OF AMMONIA. 



On Plot 2A. On Plot 4A. 



First 10 years 9-9 8'8 bushels. 



Second 10 years ... ... ... ... 13-7 n-8 



Third 10 years ... ... ... ... ir8 12-2 



Fourth 10 years ... ... ... ... 9-6 12-1 



Mean 11-2 ii'2 



The cause of this increase in the return during the later 

 years of the experiment is to be found in the impoverishment 

 of the soil where, no nitrogenous manure has been applied. 

 The soil of plots 2 o and 4 o had originally a similar composi- 

 tion to that of 2A and 4A, it is now poorer in nitrogen, and 

 the produce of the two series of plots is no longer strictly 

 comparable. This difficulty in calculating accurately the 

 increase given by the ammonia occurs more or less in all field 

 experiments, but it is especially marked in the long continued 

 barley experiments now before us. We have already remarked 

 that the spring tillage employed in barley culture determines 

 a specially rapid diminution of soil nitrogen where no nitro- 

 genous manure is applied. 



The facts just referred to only however reduce to a mode- 

 rate extent the return obtained from the use of ammonia salts 



